Industry Use of Thermal Hydraulic Codes

1
Industry Use of Thermal Hydraulic Codes

• Robert P. Martin
• Idaho National Laboratory

2
Talking Points

• Nuclear Plant Analysis Areas
• Industry design and safety processes
• Recurring and infrequent code applications
• Current limitations and challenges
• Developing Common Goals to Advance Codes
• How Industry Codes are LIKE Laboratory Codes
• How Industry Codes are UNLIKE Laboratory Codes
• Why do we need RELAP5/6/7?
• A Path Forward

3
Selected TH Codes

• Codes for Fuel Performance Analysis
• Westinghouse proprietary
• GE proprietary
• AREVA proprietary
• FRAP (PNL/INL)
• FALCON (EPRI)
• Codes for Containment Analysis
• CONTEMPT (INL)
• MELCOR (SNL)
• GOTHIC (EPRI)
• Codes for Severe Accidents
• SCDAP/RELAP (INL)
• MELCOR (SNL)
• MAAP (EPRI)

• Codes for LWR System Analysis
• RELAP5 (INL)
• TRAC (LANL)
• RAMONA (BNL)
• COBRA (PNL)
• CATHARE (CEA)
• ATHLET (GRS)
• CATHENA (AECL)
• TRACE (NRC)
• RETRAN (EPRI)
• FATHOM (ANSYS)
• Computational Fluid Dynamics
• COMMIX (ANL)
• FLUENT (ANSYS)
• STAR-CD (CD-Adapco)

4
Plant FSAR Contents
5
Nuclear Plant Analysis Areas
6
General EM Framework
BE/Scale
Baseline
SA
Perturbation
Unc
What If
7
Reoccurring Code Applications

• Frequent Uses (reload, 18 months)
• Best-estimate (BE) fuel, reactor and BOP system
  analysis
• FSAR chapters 4, 5 10
• Both conservative and BE Design-basis safety
  analysis
• FSAR chapter 4, 6 15
• Occasional Use (every 2 or more reloads)
• Design/process modification (e.g.,
• Power uprate
• Component replacement
• Setpoints verification (IC design) FSAR
  chapter 7

8
Infrequent Code Applications

• Diversity and Defense-in-depth
• BE DBA calcs to verify secondary/tertiary control
  system performance
• FSAR chapter 7
• Structural
• Combustion and SA loads water hammer Jet
  impingement loads
• FSAR chapter 3
• Equipment qualification/survivability
• FSAR chapter 3
• Severe accident/Probabilistic Risk Assessment
  (PRA)
• FSAR chapter 19
• Spent fuel pool analysis
• FSAR chapter 9
• Accident management/simulator/training
• 10 CFR 50.34 (TMI-2 rulemaking)

9
Code Application Challenges

• Recognized code limitations
• Code variability
• Solution convergence
• Limited first principles understanding of
  phenomena
• Model approximations, bifurcations and
  discontinuities
• Errors in the Equation-of-State
• Spatial resolution to refine distributed
  transport phenomena
• Incomplete phenomenological models, e.g.,
  two-phase flows
• Best-estimate plus uncertainty (BEPU) methods
  have exposed
• Need for faster calculations (more calcs
  necessary)
• Limitations in range of applicability
• Inherent code bias and large uncertainty
• Many opportunities for the User to misapply the
  code
• Multi-physics (coupled tools), e.g., Rx kinetics,
  fuel/containment

10
R5-3D Code Variability Illustration
Martin, Quantifying Code Variability for LBLOCA
with RELAP5-3D, 2001 IRUG Meeting
11
Developing Common Goals to Advance Codes

• How Industry Codes are LIKE Laboratory Codes
• Inherited code architecture
• Basic input/output format
• Governing equations and physical models
• Order-of-solution
• Bugs
• Inherited developmental assessment
• New models/capability motivated by regulatory
  initiatives
• Multi-dimensional modeling
• Multi-physics
• New experimental data (e.g., for assessments,
  improved water properties, etc.)
• Dwindling number of developers and competent
  users

12
Developing Common Goals to Advance Codes

• How Industry Codes are UNLIKE Laboratory Codes
• Certain advanced capability neglected
• RELAP5 examples multi-dimensional reactor
  kinetics, code coupling feature, FORTRAN 90/95
• Regulations and NRCs Standard Review Plan often
  restrict application of best-estimate models or
  preclude certain legacy code models (e.g.,
  Forslund-Rohsenow film boiling)
• Expanded developmental assessment in application
  areas
• LOCA Example Fuel vendors advertise 130
  benchmarks
• Accommodation for uncertainty treatment
• Integrated multi-physics capability
• Production applications are automated

Differences reflect the manner in which industry
applies these codes
13
Developing Common Goals to Advance Codes

• Industrys focus has been on methodology
• Reflecting new plants
• Reg. Guide 1.203 (Evaluation Methodology
  Development and Assessment Process, EMDAP, 2005)
• Lab focus has been on modernization,
  sustainability, and science-based
• Modernize legacy codes, e.g., FORTRAN 90/95,
  remove unacceptable models
• Sustainability seems to mean new models to
  address current LWR issues, not current analysis
  challenges
• Science-based methods -gt first principles models
  and correlations
• Industry needs to modernize codes
• to be assured that they can continue to run
• to have a reason to sponsor development and
  sustain developer competency
• LWR sustainability from labs by aligning
  capability with industry
• Common code capability
• Address current challenges

14
The TH Modeling and Simulation Mission

• Why do we need RELAP5?
• Captures TH modeling and analysis advances of
  past 45 years
• Broad constituency (government, industry,
  academia)
• Why do we need RELAP6?
• Move beyond todays code challenges and expand
  code architecture for substantial improvement
• Stimulate new TH RD and training methods
• Why do we need RELAP7?
• To advance the mission of the lab to further
  understanding of TH and, specifically, tools to
  advance the theory of basic processes and the
  design of complex energy systems using advanced
  numerical modeling and computer simulations

15
A Path Forward Summary

• Industry needs to move forward with code
  modernization
• For RELAP5-based methodologies, it might be
  easier to use current version, modify i- and
  r-level routines, and integrate unique models
• PVM/MPI message passing capability should be
  added to open TH codes to enhancement available
  from tools catering to specific code needs
• Lab needs to recognize and act on industry needs
• Integrate multi-physics capability as currently
  done in industry
• Address current problems
• Code variability
• Refine spatial resolution
• Improve important phenomenological models
• Accommodate BEPU methods
• Examine opportunities to address the User-Effect
  (e.g., automation)

16
QD Vision
SubChannel/CFD
PVM
FUEL
Containment
RELAP5-3D
MPI
MOOSE